2547: KCNT1 Round 2

Closed since over 1 year ago

Summary

• Created: December 13, 2024
• Expires: December 20, 2024 at 23:00 UTC
• Max points: 100

Description

KCNT1 is a sodium-activated potassium channel that critically influences how neurons fire. When functioning normally, KCNT1 helps maintain a healthy balance of electrical signaling in the brain. However, mutations in the KCNT1 gene can lead to a range of severe epileptic conditions such as Malignant Migrating Partial Seizures of Infancy (MMPSI).

Designing inhibitors that can precisely modulate KCNT1’s activity may offer pathways to managing or even alleviating some resulting debilitating conditions. In this puzzle series, you’ll start with a candidate molecule nestled into a KCNT1 binding pocket. Your goal is to create a potential therapeutic lead. By applying your intuition and problem-solving skills here, you could help lay the groundwork for novel treatments targeting severe pediatric epilepsies and related neurological disorders.

Top groups

• 1. Anthropic Dreams 100 pts. 23,590
• 2. VeFold 73 pts. 22,453
• 3. Contenders 52 pts. 21,759
• 4. Go Science 36 pts. 21,037
• 5. L'Alliance Francophone 24 pts. 20,621
• 6. Australia 16 pts. 20,057
• 7. Marvin's bunch 10 pts. 18,449
• 8. Team China 6 pts. 18,122
• 9. FamilyBarmettler 4 pts. 18,078
• 10. Russian team 2 pts. 17,257

Top soloists

• 1. gmn 100 pts. 23,575
• 2. meatexplosion 94 pts. 23,210
• 3. JuliaBCollet 88 pts. 22,133
• 4. BootsMcGraw 82 pts. 21,624
• 5. abiogenesis 76 pts. 21,037
• 6. Th1sN@me!sN0tAPun 71 pts. 20,947
• 7. christioanchauvin 66 pts. 20,621
• 8. georg137 61 pts. 20,494
• 9. LociOiling 56 pts. 20,425
• 10. AlkiP0Ps 52 pts. 20,057

Comments

[Sciren]

Here's an example of how to find a floating ligand. It finds the first ligand in a puzzle, which should cover small molecule design puzzles for now.

In the example, PST is a table used to the store the ligand segment and other "puzzle state" info. Some recipes might have plain global variables LIGAND and AtomsInLigand, not stored in a table.

function HasLigand () 
    local segCnt = structure.GetCount ()
    local HASLIGAND = false
    for ii = 1, structure.GetCount () do
        if structure.GetSecondaryStructure ( ii ) == "M" then
            PST.LIGAND = ii
            PST.AtomsInLigand = structure.GetAtomCount(PST.LIGAND)
            HASLIGAND = true 
            break
        end
    end
    return HASLIGAND
end

[LociOiling]

This puzzle has the ligand at segment 141 out of 293 segments.

In most recent ligand puzzles over the past several years, the ligand is always the highest segment number.

The unusual position breaks many ligand recipes, which assume the ligand must be the last segment.

Back in the middle ages, some puzzles had ligands in the middle, and some puzzles had more than one ligand. Usually multiple ligands involved metal atoms.

[LociOiling]

I'm also noticing that recipes tend to miss gains on 2547. I didn't see this on last week's KCNT1 puzzle.

A missed gain means that the recipe failed to detect an increase in score for some reason. A "Restore Credit Best" at the end of the recipe is a partial fix, but that means the recipe wasted effort working on lower-scoring poses.

I'm also seeing a common issue involving the rotamer count. Initially, the rotamer.GetCount function returns 1 on the starting pose for 2547. After working the puzzle a little, this condition normally goes away. A simple shake and wiggle is not enough however.

The recipe Lig Detector 1.1 shows both the unusual ligand location and the rotamer count issue.

[LociOiling]

I'm also noticing that recipes tend to miss gains on 2547. I didn't see this on last week's KCNT1 puzzle.

A missed gain means that the recipe failed to detect an increase in score for some reason. A "Restore Credit Best" at the end of the recipe is a partial fix, but that means the recipe wasted effort working on lower-scoring poses.

I'm also seeing a common issue involving the rotamer count. Initially, the rotamer.GetCount function returns 1 on the starting pose for 2547. After working the puzzle a little, this condition normally goes away. A simple shake and wiggle is not enough however.

The recipe Lig Detector 1.1 shows both the unusual ligand location and the rotamer count issue.